Wei-Shung Chang

A competitive evolution strategy memetic algorithm for unrelated parallel machine scheduling to minimize total weighted tardiness and flow time

This research proposes a competitive evolution strategy memetic algorithm (CESMA) to solve unrelated parallel machines scheduling problems with two minimization objectives subject to job sequence- and machine-dependent setup times. A memetic operation is regarded as a genetic operation following a local search-weighted bipartite matching algorithm (WBM). The competitive evolution strategy maintains one generational population (GP) and two external archives at each generation, one preserving efficient solutions and the other preserving inefficient solutions. At each generation, two procedures, EAMA (efficient archive memetic algorithm) and IAMA (inefficient archive memetic algorithm), are applied to compete for producing the next generation offspring. The fraction p of memetic operations assigned to EAMA varies at each generation and depends on the competition results of the last generation. An experiment is conducted to compare the performance of the CESMA against two well-known evolutionary algorithms (NSGA II and SPEA2) with WBM. The effects of incorporating the WBM into these algorithms are also investigated. In the experimental study, three instances of different problem parameters were generated using a method in the literature. The experimental results show that the CESMA excels the others in terms of several proximity measures

Hybrid Metaheuristics for the Unrelated Parallel Machine Scheduling to Minimize Makespan and Maximum Just-in-Time Deviations

This paper studies the unrelated parallel machine scheduling problem with three minimization objectives - makespan, maximum earliness, and maximum tardiness (MET- UPMSP). The last two objectives combined are related to just-in- time (JIT) performance of a solution. Three hybrid algorithms are presented to solve the MET-UPMSP: reactive GRASP with path relinking, dual-archived memetic algorithm (DAMA), and SPEA2. In order to improve the solution quality, min-max matching is included in the decoding scheme for each algorithm. An experiment is conducted to evaluate the performance of the three algorithms, using 100 (jobs) x 3 (machines) and 200 x 5 problem instances with three combinations of two due date factors - tight and range. The numerical results indicate that DAMA performs best and GRASP performs second for most problem instances in three performance metrics: HVR, GD, and Spread. The experimental results also show that incorporating min-max matching into decoding scheme significantly improves the solution quality for the two population-based algorithms. It is worth noting that the solutions produced by DAMA with matching decoding can be used as benchmark to evaluate the performance of other algorithms.

A Multi-Criteria Decision Making for the Unrelated Parallel Machines Scheduling Problem

In this paper, we propose a multi-criteria machine-schedules decision making method that can be applied to a produc-tion environment involving several unrelated parallel machines and we will focus on three objectives: minimizing makespan, total flow time, and total number of tardy jobs. The decision making method consists of three phases. In the first phase, a mathematical model of a single machine scheduling problem, of which the objective is a weighted sum of the three objectives, is constructed. Such a model will be repeatedly solved by the CPLEX in the proposed Multi-Objective Simulated Annealing (MOSA) algorithm. In the second phase, the MOSA that integrates job clustering method, job group scheduling method, and job group – machine assignment method, is employed to obtain a set of non-dominated group schedules. During this phase, CPLEX software and the bipartite weighted matching algorithm are used repeatedly as parts of the MOSA algorithm. In the last phase, the technique of data envelopment analysis is applied to determine the most preferable schedule. A practical example is then presented in order to demonstrate the applicability of the proposed decision making method.

A Dual-Archive Memetic Algorithm for Unrelated Parallel-Machine Scheduling to Minimize Total Weighted Flow Time and Weighted Tardiness

Abstract This paper presents an effective approach to solving unrelated parallel-machine scheduling problems that minimizes two aggregation objectives: total weighted flow time and total weighted tardiness. At each iteration step, the approach partitions the objective space using different weights on each objective, and applies weighted bipartite matching (WBM) to find the best neighborhood solution in each objective subspace. Three algorithms are used to assess this approach: NSGAII, SPEA2, and DAMA (dual-archive memetic algorithm). When using WBM, weighted apparent tardiness cost with setups (W-ATCS) is employed to solve single machine scheduling problems. For DAMA, two dissimilar archives are maintained at each generation: one archive preserves efficient solutions, the other preserves inefficient solutions, and the two archives compete to produce next generation offspring. An experiment was conducted to evaluate the proposed approach based on several performance metrics. The results indicate that decoding scheme using WBM will produce significantly better solutions, regardless of which algorithm is employed. The results also show that using random weights (RW) on objectives for evolution excels using fixed weights (FW). Finally, DAMA_RW outperforms all other algorithms based on the same number of calculated solutions.

Minimizing Total Weighted Tardiness and Flow Time for Unrelated Parallel Machines Scheduling with Matching-Improvement Evolutionary Algorithms

Abstract This paper addresses unrelated parallel machine scheduling problems with two minimization objectives: total weighted flow time and tardiness, and presents two hybrid methods based on (1) non-dominated sorting genetic algorithms (NSGA-II) and (2) strength Pareto evolutionary algorithm (SPEA). These algorithms were implemented in a different manner according to the following two features: (1) using random or fixed weighted sum direction search (RWSD or FWSD); (2) including or not including a bipartite weighted matching problem (BWMP). The performance of the algorithms is evaluated via two benchmark instances generated by a method in the literature. The experimental results indicate that algorithms with RWSD are superior to those with FWSD, and those including BWMP outperforms those not, in terms of proximity and spread metrics. In particular, NSGA-II with RWSD and BWMP performs best for the large size instance, whereas SPEA with RWSD and BWMP excels other algorithms in solving the medium size instance. Nevertheless, algorithms without BWMP spend much less computation time than others under the same termination criterion